In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
import seaborn as sns
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix
from sklearn.metrics import roc_auc_score
from sklearn.model_selection import cross_val_score
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import classification_report
from sklearn.metrics import accuracy_score
from sklearn.model_selection import RepeatedStratifiedKFold
from sklearn.metrics import precision_recall_curve
pd.options.display.float_format = '{:.2f}'.format
import warnings
warnings.filterwarnings('ignore')
In [2]:
data = pd.read_csv('heart.csv')
data.head()
Out[2]:
Age Sex ChestPainType RestingBP Cholesterol FastingBS RestingECG MaxHR ExerciseAngina Oldpeak ST_Slope HeartDisease
0 40 M ATA 140 289 0 Normal 172 N 0.00 Up 0
1 49 F NAP 160 180 0 Normal 156 N 1.00 Flat 1
2 37 M ATA 130 283 0 ST 98 N 0.00 Up 0
3 48 F ASY 138 214 0 Normal 108 Y 1.50 Flat 1
4 54 M NAP 150 195 0 Normal 122 N 0.00 Up 0
In [4]:
data.shape
Out[4]:
(918, 12)
In [5]:
data.columns
Out[5]:
Index(['Age', 'Sex', 'ChestPainType', 'RestingBP', 'Cholesterol', 'FastingBS',
       'RestingECG', 'MaxHR', 'ExerciseAngina', 'Oldpeak', 'ST_Slope',
       'HeartDisease'],
      dtype='object')
In [6]:
data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 918 entries, 0 to 917
Data columns (total 12 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   Age             918 non-null    int64  
 1   Sex             918 non-null    object 
 2   ChestPainType   918 non-null    object 
 3   RestingBP       918 non-null    int64  
 4   Cholesterol     918 non-null    int64  
 5   FastingBS       918 non-null    int64  
 6   RestingECG      918 non-null    object 
 7   MaxHR           918 non-null    int64  
 8   ExerciseAngina  918 non-null    object 
 9   Oldpeak         918 non-null    float64
 10  ST_Slope        918 non-null    object 
 11  HeartDisease    918 non-null    int64  
dtypes: float64(1), int64(6), object(5)
memory usage: 86.2+ KB
In [7]:
data.describe().T
Out[7]:
count mean std min 25% 50% 75% max
Age 918.00 53.51 9.43 28.00 47.00 54.00 60.00 77.00
RestingBP 918.00 132.40 18.51 0.00 120.00 130.00 140.00 200.00
Cholesterol 918.00 198.80 109.38 0.00 173.25 223.00 267.00 603.00
FastingBS 918.00 0.23 0.42 0.00 0.00 0.00 0.00 1.00
MaxHR 918.00 136.81 25.46 60.00 120.00 138.00 156.00 202.00
Oldpeak 918.00 0.89 1.07 -2.60 0.00 0.60 1.50 6.20
HeartDisease 918.00 0.55 0.50 0.00 0.00 1.00 1.00 1.00
In [8]:
yes = data[data['HeartDisease'] == 1].describe().T
no = data[data['HeartDisease'] == 0].describe().T
colors = ['#F93822','#FDD20E']

fig,ax = plt.subplots(nrows = 1,ncols = 2,figsize = (5,5))
plt.subplot(1,2,1)
sns.heatmap(yes[['mean']],annot = True,cmap = colors,linewidths = 0.4,linecolor = 'black',cbar = False,fmt = '.2f',)
plt.title('Heart Disease');

plt.subplot(1,2,2)
sns.heatmap(no[['mean']],annot = True,cmap = colors,linewidths = 0.4,linecolor = 'black',cbar = False,fmt = '.2f')
plt.title('No Heart Disease');

fig.tight_layout(pad = 2)
In [11]:
col = list(data.columns)
categorical_features = []
numerical_features = []
for i in col:
    if len(data[i].unique()) > 6:
        numerical_features.append(i)
    else:
        categorical_features.append(i)

print('Categorical Features :',*categorical_features)
print('Numerical Features :',*numerical_features)

Categorical Features : Sex ChestPainType FastingBS RestingECG ExerciseAngina ST_Slope HeartDisease
Numerical Features : Age RestingBP Cholesterol MaxHR Oldpeak
In [12]:
col = list(data.columns)
categorical_features = []
numerical_features = []
for i in col:
    if len(data[i].unique()) > 6:
        numerical_features.append(i)
    else:
        categorical_features.append(i)

print('Categorical Features :',*categorical_features)
print('Numerical Features :',*numerical_features)

Categorical Features : Sex ChestPainType FastingBS RestingECG ExerciseAngina ST_Slope HeartDisease
Numerical Features : Age RestingBP Cholesterol MaxHR Oldpeak
In [50]:
fig, ax = plt.subplots(nrows = 2,ncols = 2,figsize = (10,9.75))
for i in range(len(numerical_features) - 1):
    plt.subplot(2,2,i+1)
    sns.distplot(data[numerical_features[i]],color = colors[0])
    title = 'Distribution : ' + numerical_features[i]
    plt.title(title)
plt.show()

plt.figure(figsize = (4.75,4.55))
sns.distplot(df1[numerical_features[len(numerical_features) - 1]],kde_kws = {'bw' : 1},color = colors[0])
title = 'Distribution : ' + numerical_features[len(numerical_features) - 1]
plt.title(title);
In [57]:
l = list(data['HeartDisease'].value_counts())
circle = [l[1] / sum(l) * 100,l[0] / sum(l) * 100]

plt.subplot(1,2,1)
plt.pie(circle,labels = ['No Heart Disease','Heart Disease'],autopct='%1.1f%%',startangle = 90,explode = (0.1,0),colors = colors,
        wedgeprops = {'edgecolor' : 'black','linewidth': 1,'antialiased' : True})
plt.title('Heart Disease %');
In [55]:
fig, ax = plt.subplots(nrows=3, ncols=2, figsize=(10, 15))

for i in range(len(categorical_features) - 1):
    plt.subplot(3, 2, i + 1)
    ax = sns.countplot(x=categorical_features[i], data=data, hue="HeartDisease", palette=colors, edgecolor='black')
    
    for rect in ax.patches:
        ax.text(rect.get_x() + rect.get_width() / 2, rect.get_height() + 2, rect.get_height(), horizontalalignment='center', fontsize=11)
    
    title = categorical_features[i] + ' vs HeartDisease'
    plt.legend(['No Heart Disease', 'Heart Disease'])
    plt.title(title)

plt.tight_layout()
plt.show()
In [58]:
sex = data[data['HeartDisease'] == 1]['Sex'].value_counts()
sex = [sex[0] / sum(sex) * 100, sex[1] / sum(sex) * 100]

cp = data[data['HeartDisease'] == 1]['ChestPainType'].value_counts()
cp = [cp[0] / sum(cp) * 100,cp[1] / sum(cp) * 100,cp[2] / sum(cp) * 100,cp[3] / sum(cp) * 100]

fbs = data[data['HeartDisease'] == 1]['FastingBS'].value_counts()
fbs = [fbs[0] / sum(fbs) * 100,fbs[1] / sum(fbs) * 100]

restecg = data[data['HeartDisease'] == 1]['RestingECG'].value_counts()
restecg = [restecg[0] / sum(restecg) * 100,restecg[1] / sum(restecg) * 100,restecg[2] / sum(restecg) * 100]

exang = data[data['HeartDisease'] == 1]['ExerciseAngina'].value_counts()
exang = [exang[0] / sum(exang) * 100,exang[1] / sum(exang) * 100]

slope = data[data['HeartDisease'] == 1]['ST_Slope'].value_counts()
slope = [slope[0] / sum(slope) * 100,slope[1] / sum(slope) * 100,slope[2] / sum(slope) * 100]
In [59]:
ax,fig = plt.subplots(nrows = 4,ncols = 2,figsize = (15,15))

plt.subplot(3,2,1)
plt.pie(sex,labels = ['Male','Female'],autopct='%1.1f%%',startangle = 90,explode = (0.1,0),colors = colors,
        wedgeprops = {'edgecolor' : 'black','linewidth': 1,'antialiased' : True})
plt.title('Sex');

plt.subplot(3,2,2)
plt.pie(cp,labels = ['ASY', 'NAP', 'ATA', 'TA'],autopct='%1.1f%%',startangle = 90,explode = (0,0.1,0.1,0.1),
        wedgeprops = {'edgecolor' : 'black','linewidth': 1,'antialiased' : True})
plt.title('ChestPainType');

plt.subplot(3,2,3)
plt.pie(fbs,labels = ['FBS < 120 mg/dl','FBS > 120 mg/dl'],autopct='%1.1f%%',startangle = 90,explode = (0.1,0),colors = colors,
        wedgeprops = {'edgecolor' : 'black','linewidth': 1,'antialiased' : True})
plt.title('FastingBS');

plt.subplot(3,2,4)
plt.pie(restecg,labels = ['Normal','ST','LVH'],autopct='%1.1f%%',startangle = 90,explode = (0,0.1,0.1),
        wedgeprops = {'edgecolor' : 'black','linewidth': 1,'antialiased' : True})
plt.title('RestingECG');

plt.subplot(3,2,5)
plt.pie(exang,labels = ['Angina','No Angina'],autopct='%1.1f%%',startangle = 90,explode = (0.1,0),colors = colors,
        wedgeprops = {'edgecolor' : 'black','linewidth': 1,'antialiased' : True})
plt.title('ExerciseAngina');

plt.subplot(3,2,6)
plt.pie(slope,labels = ['Flat','Up','Down'],autopct='%1.1f%%',startangle = 90,explode = (0,0.1,0.1),
        wedgeprops = {'edgecolor' : 'black','linewidth': 1,'antialiased' : True})
plt.title('ST_Slope');
In [65]:
fig, ax = plt.subplots(nrows=5, ncols=1, figsize=(15, 30))

for i in range(len(numerical_features)):
    plt.subplot(5, 1, i + 1)
    sns.countplot(x=numerical_features[i], data=data, hue="HeartDisease", palette=colors, edgecolor='black')
    
    title = numerical_features[i] + ' vs Heart Disease'
    plt.legend(['No Heart Disease', 'Heart Disease'])
    plt.title(title)

plt.tight_layout()
plt.show()
In [66]:
data['RestingBP_Group'] = [ int(i / 5) for i in data['RestingBP']]
data['Cholesterol_Group'] = [ int(i / 10) for i in data['Cholesterol']]
data['MaxHR_Group'] = [ int(i / 5) for i in data['MaxHR']]
data['Oldpeak_Group'] = [ int( (i*10) / 5) for i in data['Oldpeak']]
In [67]:
fig, ax = plt.subplots(nrows = 4,ncols = 1,figsize = (10,25))
group_numerical_features = [i + '_Group' for i in numerical_features[1:]]
for i in range(len(group_numerical_features)):
    plt.subplot(4,1,i+1)
    sns.countplot(x=group_numerical_features[i],data = data,hue = "HeartDisease",palette = colors, edgecolor = 'black')
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = group_numerical_features[i] + ' vs Heart Disease'
    plt.title(title);
In [68]:
fig,ax = plt.subplots(nrows = 1,ncols = 3,figsize = (15,5))
for i in range(3):
    plt.subplot(1,3,i+1)
    sns.stripplot(x = 'Sex',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs Sex' 
    plt.title(title);

fig,ax = plt.subplots(nrows = 1,ncols = 2,figsize = (15,5))
for i in [-1,-2]:
    plt.subplot(1,2,-i)
    sns.stripplot(x = 'Sex',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs Sex' 
    plt.title(title);
In [69]:
fig,ax = plt.subplots(nrows = 1,ncols = 3,figsize = (15,5))
for i in range(3):
    plt.subplot(1,3,i+1)
    sns.stripplot(x = 'ChestPainType',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs ChestPainType'
    plt.title(title);

fig,ax = plt.subplots(nrows = 1,ncols = 2,figsize = (15,5))
for i in [-1,-2]:
    plt.subplot(1,2,-i)
    sns.stripplot(x = 'ChestPainType',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs ChestPainType' 
    plt.title(title);
In [70]:
fig,ax = plt.subplots(nrows = 1,ncols = 3,figsize = (15,5))
for i in range(3):
    plt.subplot(1,3,i+1)
    sns.stripplot(x = 'FastingBS',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs Fasting Blood Sugar' 
    plt.title(title);

fig,ax = plt.subplots(nrows = 1,ncols = 2,figsize = (15,5))
for i in [-1,-2]:
    plt.subplot(1,2,-i)
    sns.stripplot(x = 'FastingBS',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs Fasting Blood Sugar' 
    plt.title(title);
In [71]:
fig,ax = plt.subplots(nrows = 1,ncols = 3,figsize = (15,5))
for i in range(3):
    plt.subplot(1,3,i+1)
    sns.stripplot(x = 'RestingECG',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs RestingECG' 
    plt.title(title);

fig,ax = plt.subplots(nrows = 1,ncols = 2,figsize = (15,5))
for i in [-1,-2]:
    plt.subplot(1,2,-i)
    sns.stripplot(x = 'RestingECG',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs RestingECG'
    plt.title(title);
In [72]:
fig,ax = plt.subplots(nrows = 1,ncols = 3,figsize = (15,5))
for i in range(3):
    plt.subplot(1,3,i+1)
    sns.stripplot(x = 'ExerciseAngina',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs ExerciseAngina'
    plt.title(title);

fig,ax = plt.subplots(nrows = 1,ncols = 2,figsize = (15,5))
for i in [-1,-2]:
    plt.subplot(1,2,-i)
    sns.stripplot(x = 'ExerciseAngina',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs ExerciseAngina' 
    plt.title(title);
In [73]:
fig,ax = plt.subplots(nrows = 1,ncols = 3,figsize = (15,5))
for i in range(3):
    plt.subplot(1,3,i+1)
    sns.stripplot(x = 'ST_Slope',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs ST_Slope' 
    plt.title(title);

fig,ax = plt.subplots(nrows = 1,ncols = 2,figsize = (15,5))
for i in [-1,-2]:
    plt.subplot(1,2,-i)
    sns.stripplot(x = 'ST_Slope',y = numerical_features[i],data = data,hue = 'HeartDisease',palette = colors);
    plt.legend(['No Heart Disease', 'Heart Disease'])
    title = numerical_features[i] + ' vs ST_Slope'
    plt.title(title);
In [74]:
a = 0
fig,ax = plt.subplots(nrows = 5,ncols = 2,figsize = (15,25))
for i in range(len(numerical_features)):
    for j in range(len(numerical_features)):
        if i != j and j > i:
            a += 1
            plt.subplot(5,2,a)
            sns.scatterplot(x = numerical_features[i],y = numerical_features[j],data = data,hue = 'HeartDisease',palette = colors, edgecolor = 'black');
            plt.legend(['No Heart Disease', 'Heart Disease'])
            title = numerical_features[i] + ' vs ' + numerical_features[j]
            plt.title(title)
In [75]:
from sklearn.preprocessing import MinMaxScaler,StandardScaler
mms = MinMaxScaler() # Normalization
ss = StandardScaler() # Standardization

df1['Oldpeak'] = mms.fit_transform(df1[['Oldpeak']])
df1['Age'] = ss.fit_transform(df1[['Age']])
df1['RestingBP'] = ss.fit_transform(df1[['RestingBP']])
df1['Cholesterol'] = ss.fit_transform(df1[['Cholesterol']])
df1['MaxHR'] = ss.fit_transform(df1[['MaxHR']])
df1.head()
Out[75]:
Age Sex ChestPainType RestingBP Cholesterol FastingBS RestingECG MaxHR ExerciseAngina Oldpeak ST_Slope HeartDisease
0 -1.43 1 1 0.41 0.83 0 1 1.38 0 0.30 2 0
1 -0.48 0 2 1.49 -0.17 0 1 0.75 0 0.41 1 1
2 -1.75 1 1 -0.13 0.77 0 2 -1.53 0 0.30 2 0
3 -0.58 0 0 0.30 0.14 0 1 -1.13 1 0.47 1 1
4 0.05 1 2 0.95 -0.03 0 1 -0.58 0 0.30 2 0
In [76]:
plt.figure(figsize = (20,5))
sns.heatmap(df1.corr(),cmap = colors,annot = True);
In [77]:
corr = df1.corrwith(df1['HeartDisease']).sort_values(ascending = False).to_frame()
corr.columns = ['Correlations']
plt.subplots(figsize = (5,5))
sns.heatmap(corr,annot = True,cmap = colors,linewidths = 0.4,linecolor = 'black');
plt.title('Correlation w.r.t HeartDisease');
In [78]:
from sklearn.feature_selection import SelectKBest
from sklearn.feature_selection import chi2
In [79]:
features = df1.loc[:,categorical_features[:-1]]
target = df1.loc[:,categorical_features[-1]]

best_features = SelectKBest(score_func = chi2,k = 'all')
fit = best_features.fit(features,target)

featureScores = pd.DataFrame(data = fit.scores_,index = list(features.columns),columns = ['Chi Squared Score']) 

plt.subplots(figsize = (5,5))
sns.heatmap(featureScores.sort_values(ascending = False,by = 'Chi Squared Score'),annot = True,cmap = colors,linewidths = 0.4,linecolor = 'black',fmt = '.2f');
plt.title('Selection of Categorical Features');
In [80]:
from sklearn.feature_selection import f_classif

features = df1.loc[:,numerical_features]
target = df1.loc[:,categorical_features[-1]]

best_features = SelectKBest(score_func = f_classif,k = 'all')
fit = best_features.fit(features,target)

featureScores = pd.DataFrame(data = fit.scores_,index = list(features.columns),columns = ['ANOVA Score']) 

plt.subplots(figsize = (5,5))
sns.heatmap(featureScores.sort_values(ascending = False,by = 'ANOVA Score'),annot = True,cmap = colors,linewidths = 0.4,linecolor = 'black',fmt = '.2f');
plt.title('Selection of Numerical Features');
In [82]:
pip install --upgrade scikit-learn

Requirement already satisfied: scikit-learn in c:\users\lenovo\anaconda3\lib\site-packages (1.3.0)
Requirement already satisfied: numpy>=1.17.3 in c:\users\lenovo\anaconda3\lib\site-packages (from scikit-learn) (1.24.3)
Requirement already satisfied: scipy>=1.5.0 in c:\users\lenovo\anaconda3\lib\site-packages (from scikit-learn) (1.10.1)
Requirement already satisfied: joblib>=1.1.1 in c:\users\lenovo\anaconda3\lib\site-packages (from scikit-learn) (1.2.0)
Requirement already satisfied: threadpoolctl>=2.0.0 in c:\users\lenovo\anaconda3\lib\site-packages (from scikit-learn) (2.2.0)
Note: you may need to restart the kernel to use updated packages.
In [96]:
features = df1[df1.columns.drop(['HeartDisease','RestingBP','RestingECG'])].values
target = df1['HeartDisease'].values
x_train, x_test, y_train, y_test = train_test_split(features, target, test_size = 0.20, random_state = 2)
In [115]:
def model(classifier):
    
    classifier.fit(x_train,y_train)
    prediction = classifier.predict(x_test)
    cv = RepeatedStratifiedKFold(n_splits = 10,n_repeats = 3,random_state = 1)
    print("Accuracy : ",'{0:.2%}'.format(accuracy_score(y_test,prediction)))
    print("Cross Validation Score : ",'{0:.2%}'.format(cross_val_score(classifier,x_train,y_train,cv = cv,scoring = 'roc_auc').mean()))
    print("ROC_AUC Score : ",'{0:.2%}'.format(roc_auc_score(y_test,prediction)))

def model_evaluation(classifier):
    
    # Confusion Matrix
    cm = confusion_matrix(y_test,classifier.predict(x_test))
    names = ['True Neg','False Pos','False Neg','True Pos']
    counts = [value for value in cm.flatten()]
    percentages = ['{0:.2%}'.format(value) for value in cm.flatten()/np.sum(cm)]
    labels = [f'{v1}\n{v2}\n{v3}' for v1, v2, v3 in zip(names,counts,percentages)]
    labels = np.asarray(labels).reshape(2,2)
    
    # Classification Report
    print(classification_report(y_test,classifier.predict(x_test)))
In [108]:
classifier_lr = LogisticRegression(random_state = 0,C=10,penalty= 'l2')
In [116]:
model(classifier_lr)

Accuracy :  87.50%
Cross Validation Score :  91.12%
ROC_AUC Score :  87.43%
In [117]:
model_evaluation(classifier_lr)

              precision    recall  f1-score   support

           0       0.88      0.85      0.87        89
           1       0.87      0.89      0.88        95

    accuracy                           0.88       184
   macro avg       0.88      0.87      0.87       184
weighted avg       0.88      0.88      0.87       184
In [118]:
from sklearn.svm import SVC
In [119]:
classifier_svc = SVC(kernel = 'linear',C = 0.1)
In [120]:
model(classifier_svc)

Accuracy :  87.50%
Cross Validation Score :  90.53%
ROC_AUC Score :  87.43%
In [121]:
model_evaluation(classifier_svc)

              precision    recall  f1-score   support

           0       0.88      0.85      0.87        89
           1       0.87      0.89      0.88        95

    accuracy                           0.88       184
   macro avg       0.88      0.87      0.87       184
weighted avg       0.88      0.88      0.87       184
In [122]:
from sklearn.tree import DecisionTreeClassifier
In [123]:
classifier_dt = DecisionTreeClassifier(random_state = 1000,max_depth = 4,min_samples_leaf = 1)
In [124]:
model(classifier_dt)

Accuracy :  84.78%
Cross Validation Score :  89.09%
ROC_AUC Score :  84.62%
In [125]:
model_evaluation(classifier_dt)

              precision    recall  f1-score   support

           0       0.88      0.80      0.84        89
           1       0.83      0.89      0.86        95

    accuracy                           0.85       184
   macro avg       0.85      0.85      0.85       184
weighted avg       0.85      0.85      0.85       184